The sensorless predictive control scheme based on fusion position estimation technology is proposed for the low-speed heavy-load case of permanent magnet synchronous machines (PMSMs). According to the motor voltage equation, the optimization-based method can estimate the rotor information over the full speed range. However, core saturation occurs when the PMSM is heavily loaded at low-speed conditions, resulting in the failure of the optimization-based method. To address this, this article proposes an extended Kalman filter (EKF) observer on the basis of the mechanical dynamic model. The EKF observer can integrate the estimation of the mechanical model based with that of the electrical model, which can effectively filter out noise terms. Then, the estimation error is compensated by introducing an adaptive coefficient to further improve the estimation effect. Meanwhile, a robust generalized predictive controller (RGPC) is designed for PMSM drives, where the feedback information is derived from the fusion estimates. The proposed sensorless predictive control with the fusion position estimation technology is simulated and experimentally verified on a heavy-load pendulum servo bench.
Sensorless Predictive Control With Fusion Position Estimation Technology for PMSM Drives
IEEE Transactions on Transportation Electrification ; 10 , 3 ; 5026-5036
2024-09-01
3248633 byte
Article (Journal)
Electronic Resource
English
FPGA-based high performance sensorless control for PMSM drives
TIBKAT | 2014
|PMSM Sensorless Control Based on Position Estimation Correction Using All‐Pass Filters
Online Contents | 2017
|